 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
When I rendered this file, I got this:
http://news.povray.org/povray.binaries.images/thread/%3Cweb.551e92b74514bfb213709ea20%40news.povray.org%3E/
And here is a discussion several days ago, which may be a similar phenomenon.
http://news.povray.org/povray.bugreports/thread/%3Cweb.5517c69b250fa02e5189b500%40news.povray.org%3E/
Post a reply to this message
Attachments:
Download 'bugreport.pov.txt' (2 KB)
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> When I rendered this file, I got this:
>
http://news.povray.org/povray.binaries.images/thread/%3Cweb.551e92b74514bfb213709ea20%40news.povray.org%3E/
>
>
>
> And here is a discussion several days ago, which may be a similar phenomenon.
>
http://news.povray.org/povray.bugreports/thread/%3Cweb.5517c69b250fa02e5189b500%40news.povray.org%3E/
>
It looks like it could be a precision/accuracy isue as multiplying the
dimentions by 50 makes the artefact disapear. You may hit the limits of
double precision FP numbers.
Scaling up, even by 1000, DON'T improve the end result. In fact, when
scalling up, I get a full black circle, not just parts of it.
Alain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Le 04/04/2015 04:35, Alain a écrit :
> Le 15-04-03 09:25, And a écrit :
>> When I rendered this file, I got this:
>>
http://news.povray.org/povray.binaries.images/thread/%3Cweb.551e92b74514bfb213709ea20%40news.povray.org%3E/
>>
>>
>>
>>
>> And here is a discussion several days ago, which may be a similar
>> phenomenon.
>>
http://news.povray.org/povray.bugreports/thread/%3Cweb.5517c69b250fa02e5189b500%40news.povray.org%3E/
>>
>>
>
> It looks like it could be a precision/accuracy isue as multiplying the
> dimentions by 50 makes the artefact disapear. You may hit the limits of
> double precision FP numbers.
>
> Scaling up, even by 1000, DON'T improve the end result. In fact, when
> scalling up, I get a full black circle, not just parts of it.
>
>
>
> Alain
I just scaled it by 10 and it went away:
torus{
-0.1247, 0.2357
rotate <-90,0,0>
scale 1/10
}
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thank you.
I explain my intend of doing it. I made a macro that is used for chamfering a
cylinder and it is necessary to take the torus's inner surface in some
conditions. So I intersect sphere and torus with negative major_radius and
sometimes found these black lines.
Maybe the same demand can be done by using polynomial's torus and I would like
to try some day. Because I thought the black line may be produced by the
bounding cylinder of the torus object that the document describes.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 05.04.2015 um 06:25 schrieb And:
> Thank you.
> I explain my intend of doing it. I made a macro that is used for chamfering a
> cylinder and it is necessary to take the torus's inner surface in some
> conditions. So I intersect sphere and torus with negative major_radius and
> sometimes found these black lines.
Have you tried using a positive major_radius and the "inverse" keyword
instead?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
clipka <ano### [at] anonymous org> wrote:
> Am 05.04.2015 um 06:25 schrieb And:
> > Thank you.
> > I explain my intend of doing it. I made a macro that is used for chamfering a
> > cylinder and it is necessary to take the torus's inner surface in some
> > conditions. So I intersect sphere and torus with negative major_radius and
> > sometimes found these black lines.
>
> Have you tried using a positive major_radius and the "inverse" keyword
> instead?
Oh, I will try later. I just found that the torus object's overlay area is not
inside the object.
And when I use polynomial object directly it shows strong precision problem
because my torus is too small.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Le_Forgeron <jgr### [at] freefr> wrote:
> Le 04/04/2015 04:35, Alain a écrit :
> >
> > It looks like it could be a precision/accuracy isue as multiplying the
> > dimentions by 50 makes the artefact disapear. You may hit the limits of
> > double precision FP numbers.
> >
> > Scaling up, even by 1000, DON'T improve the end result. In fact, when
> > scalling up, I get a full black circle, not just parts of it.
> >
> >
> >
> > Alain
>
> I just scaled it by 10 and it went away:
>
> torus{
> -0.1247, 0.2357
> rotate <-90,0,0>
> scale 1/10
> }
Sometimes it disappear, doesn't have to be a scaling problem.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"And" <49341109@ntnu.edu.tw> wrote:
> Oh, I will try later. I just found that the torus object's overlay area is not
> inside the object.
> And when I use polynomial object directly it shows strong precision problem
> because my torus is too small.
I said it shows strong precision/accuracy problem. It happened when I tested it
with inside() function. But when I test it with inverse and intersection this
time, it overturn.
This is an (inversed) polynomial torus intersect a sphere and the torus object
version.
http://news.povray.org/povray.binaries.images/thread/%3Cweb.552405c288bb499a123791000@news.povray.org%3E
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
